|
|
[ tidy-Bugs-1264455 ] Multiple URLs in profile incorrectly modified: msg#00079web.html-tidy.tracker
Bugs item #1264455, was opened at 2005-08-20 01:08 Message generated for change (Comment added) made by hoehrmann You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=390963&aid=1264455&group_id=27659 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: HTML/XML/XHTML Parser Group: Current - all platforms Status: Open Resolution: Wont Fix Priority: 5 Submitted By: Klaus Johannes Rusch (krusch) Assigned to: Björn Höhrmann (hoehrmann) Summary: Multiple URLs in profile incorrectly modified Initial Comment: Multiple URLs in the profile attribute of a head element are incorrectly converted into a single URL with escaped whitespace, for example <head profile="http://gmpg.org/xfn/11 http://dublincore.org/documents/dcq-html/";> becomes <head profile="http://gmpg.org/xfn/11%20http://dublincore.org/documents/dcq-html/";> and a warning is issued: line 2 column 1 - Warning: <head> escaping malformed URI reference HTML 4.01 reference: http://www.w3.org/TR/REC-html40/struct/global.html#h-7.4.1 profile = uri [CT] This attribute specifies the location of one or more meta data profiles, separated by white space. For future extensions, user agents should consider the value to be a list even though this specification only considers the first URI to be significant. ---------------------------------------------------------------------- >Comment By: Björn Höhrmann (hoehrmann) Date: 2005-08-21 01:43 Message: Logged In: YES user_id=188003 Well, I don't think changing it to profile="a" helps here, we'd still get bug reports for considering profile="a b" beeing considered an error, and if we change Tidy to allow it we would get bug reports for not considering an error. ---------------------------------------------------------------------- Comment By: Klaus Johannes Rusch (krusch) Date: 2005-08-21 01:17 Message: Logged In: YES user_id=365576 Input (valid according to the wording of the HTML 4.01 specification but not matching the profile=uri pattern) <head profile="A B"> Intended interpretation according to the wording of the HTML 4.01 specification (only the first value is honored): <head profile="A"> Yes you can find arguments for both <head profile="A"> and <head profile="A B"> in the HTML specification, which obviously is not consistent here and it is unlikely that an agreement will be reached. With the interpretation instructions that only the first value is currently honored and additional values are reserved for future use, dropping the extra URIs would be a reasonable though and not change the behaviour of a conforming user agent. <head profile="A%20B"> neither matches the intent of the document author, nor the interpretation instructions in the specification. ---------------------------------------------------------------------- Comment By: Björn Höhrmann (hoehrmann) Date: 2005-08-20 20:06 Message: Logged In: YES user_id=188003 Well, this specific issue came up many many times in the past, my conclusion is that we don't do anything about it until the HTML WGs gets around to clarify their specs. You can easily avoid Tidy's behavior by using the --fix-uri option. ---------------------------------------------------------------------- Comment By: Klaus Johannes Rusch (krusch) Date: 2005-08-20 16:06 Message: Logged In: YES user_id=365576 The description does indicate though that while only one URI is currently considered significant more than one can be provided. Stripping additional URLs might be reasonable since that is what browsers are supposed to do as well, and what the DTD and later specifications support as well, converting "url1 url2" to "url1%20url2" is not since the result is a URL that is formally correct but no longer valid. With the restriction to one and the lack of a namespace-like mapping of eta elements to profile definition the profile attribute looks pretty useless anyway but at least tidy should not break the profile definition. ---------------------------------------------------------------------- Comment By: Björn Höhrmann (hoehrmann) Date: 2005-08-20 01:15 Message: Logged In: YES user_id=188003 It says "profile = uri" and that's what's been implemented in Tidy ever since. This is a known problem with the HTML and XHTML specifications. I'm happy to change Tidy's behavior once the W3C clarifies this issue. I think it's pretty clear that in HTML 4.01 and document types that build on it only a single URI is allowed and the above just describes error handling behavior. This is quite evident from even recent specifications like the M12N in XML Schema where the anyURI type is used here instead of list types. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=390963&aid=1264455&group_id=27659 ------------------------------------------------------- SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf |
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| Previous by Date: | [ tidy-Bugs-1264455 ] Multiple URLs in profile incorrectly modified: 00079, SourceForge.net |
|---|---|
| Next by Date: | [ tidy-Bugs-1264455 ] Multiple URLs in profile incorrectly modified: 00079, SourceForge.net |
| Previous by Thread: | [ tidy-Bugs-1264455 ] Multiple URLs in profile incorrectly modifiedi: 00079, SourceForge.net |
| Next by Thread: | [ tidy-Bugs-1264455 ] Multiple URLs in profile incorrectly modified: 00079, SourceForge.net |
| Indexes: | [Date] [Thread] [Top] [All Lists] |
| News | FAQ | advertise |